A Step-by-Step Guide to Exploratory Factor Analysis with SPSS

education-and-formation

• researchers tend to receive little formal training in EFA and habitually rely on suboptimal methods, modeling questionable EFA from poor-quality reports

  SSGE-RC-007

• justify the model, detail extraction method, report communalities, describe a priori criteria — multiple criteria within a model testing approach

  VLA-C9786FD7-RC-004

• conduct EFAs with and without outlier data to verify results are robust — sensitivity analysis is incumbent upon the researcher

  SSGE-RC-032

mind-and-cognition

• there is no guarantee that empirical guidelines are correct; multiple criteria including relevant theory and previous research must be used

  VLA-C62F8A01-RC-006

• parallel analysis involves comparing eigenvalues from real data against mean eigenvalues from random data; only factors above the random mean should be retained

  SSGE-RC-045

• after oblique rotation, pattern coefficients are regression-like weights; structure coefficients reflect simple correlation — it is no longer sufficient to say 'factor loadings'

  VLA-B123B705-RC-001